John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
as usual over 1/2 century old Bode Sensitivity Analysis gives good answers

... I am still looking for someone to tell me the good, bad or 'ugly' facts about the AES paper that I put up. Anybody read it? Any comments?
yes I read it - as usual over 1/2 century old Bode Sensitivity Analysis gives good answers to the 'memory problems'

basically the only places to worry are input diff pair (bootstrap cascode them) and feedback components (low TC, adequate power rating, any feedback C as least as good as NP0/C0G) in a otherwise competent high loop gain (global) negative feedback amplifier

and of course anything (else) 'outside' the feedback loop like coupling caps

'otherwise competent' only requires any memory errors like thermal bias modulation in audio power amplifier small signal stages don't result in deadbands or jump discontinuities - modulations of loop gain of factors of even a decade need not have much observable effect where loop gain/feedback is already near 100 dB - which is easy given the low frequencies of thermal effects

poorly designed Class AB could be moved by thermal effects into significantly more distorting operating points – but even Doug Self playing with his sometimes lowish 'optimum AB ” output Q bias current, while requiring more attention to thermal compensation, has been tested for 'thermal effects', (negative) results published in his books
 
Last edited:
The biggest problem I know of in she told industry is how bad the recordings and masterings etc are. They've lost so much over the years it's disturbing. But I don't expect many to understand since they think MoFi album's sound good... any of us trying to make electronics have to question if we're making something good, or just inventing new turd polishers.

By the way I have finally recovered my data on some DBT's that were pinnacle to recent advancements I've had, as I know many of you where curious. Here's a link for them.
 
The biggest problem I know of in she told industry is how bad the recordings and masterings etc are.

I have to thank you for bringing Arthur Salvatore to my attention. I just love seeing such a different viewpoint on the electronic reproduction while the appreciation of the music is just in the same vein. This tolerance of ambiguity is so missing in the conversations here.
 
Member
Joined 2014
Paid Member
Trust me windows phone is worse. My sister in law texted me the other day to say she had test driven an audi TT an liked it. What I received was 'I like T**ties. Cue enormous embarrassment from her. Mind you microsoft have a history of this. Word 5.5 for DOS couldn't handle the place 'Fucino', which was a shame as it was the main ESA ground control station at the time and was mentioned in a lot of documents we wrote.
 
Regarding fMRI, I would agree that it is a rather coarse tool. But, it is considered noninvasive, so it's something that's okay to use fairly freely in research. Other imaging modalities such as PET can show things like glucose uptake rather than just changes in blood profusion, but it means exposing patients to ionizing radiation, so it can only be used on people who happen to otherwise need PET scans for diagnostic purposes.

Still, in research people often tend to want to see some kind of brain activity that can be correlated with various experimental stimuli.

For our purposes, it may be that blood profusion changes in the brain do differ in subjects who actually recognize details of sounds in sighted listening and those who only think they do. If so, then for those that do appear to recognize details in sighted listening, do they react differently in unsighted listening to the same stimulus? Something like that. Maybe fMRI would show something and maybe nothing. I don't know. Maybe EKG would show something, but again, I don't know. I would be interested to find out though.

You see, I suspect there may be some validity to sighted listening sometimes, even though probably not most of the time. It would be nice to be able to measure something, if possible, that would help differentiate the two conditions to an external observer. There may be some other way to do it as well, or then again maybe not.
 
DBT != ABX. Sitting down and getting comfortable with two components can still be done totally double blind.

Agreed.

I don't know exactly what all the problems were in cases where people objected to various ways of being tested. It could be simple things like there wasn't enough time to get comfortable, the playback system including transducers wasn't good enough (even if the equipment specs seemed like they should be adequate), etc.

Double blind and/or ABX ought not matter, or so my intuition tells me. At least if everything else is right. If there is some problem with any reasonable test methodology, somebody should figure out what it is, and then fix it. Or figure out an alternate methodology offering equally good experimental reliability.

In addition, we probably can agree that some people who think they can hear small, subtle details, actually may not be very good at it. Personally, I always get suspicious when people talk about things like "depth of sound field," etc., because I don't know what that means, or how to translate it to more well-defined technical terminology.
 
Trust me windows phone is worse. My sister in law texted me the other day to say she had test driven an audi TT an liked it. What I received was 'I like T**ties. Cue enormous embarrassment from her. Mind you microsoft have a history of this. Word 5.5 for DOS couldn't handle the place 'Fucino', which was a shame as it was the main ESA ground control station at the time and was mentioned in a lot of documents we wrote.

Actually, the speech recognition of Windows Phone 8.1 and later are quite good. I have used all 3 phones (iOS, Android, WP) for development purposes and I would have to say that Cortana on a Lumia 950 is better than the Google Assistant on my Nexus 6P for accuracy. I could never use WP as a daily phone though as it's basically been abandoned, even by Microsoft.

All of these speech recognition technologies rely on cloud processing so if you have poor cellular signal they will all give you garbage back.

The part of the statement you need to question is that your sister liked the Audi TT :p
 
About John Curl post5091602

Hi John , hi all :)

I just went back 20 years to find a paper from the 100th AES Convention that somehow I had overlooked that appears interesting and important. Yet I am pretty sure that this paper has been debated by many of my critics here, and dismissed. Am I right?
The paper is: 'Measurement of a Neglected Circuit Characteristic' by Gerard Perrot.
This paper appears to parallel much of our work over the last 40 years when we are TRYING to improve audio, rather than dismiss it as easy and already done. I'm glad that somebody I don't know is still trying.


Maybe you don’t remember it but I told you (on the Blowtorch thread) as well as to Charles Hansen and Yoko Homo about 10 years ago about the findings of this guy, Gerard Perrot, I knew him. He died on the road, he was a biking and a car hit him. He use to write in a French audiophile magazine, after his findings on “memory distortion he founded a company named Lavardin with an associated, this company still exist and is now helded by his son and probably the associated, I don’t remember his name.

Here some links to patents and papers.

Lavardin Technologies audio systems

https://www.google.com/patents/US5512858

https://www.google.com/patents/US5635874

index

Search amplification => Héphaïstos
 

Attachments

  • Memory_Preamp.JPG
    Memory_Preamp.JPG
    29 KB · Views: 228
www.hifisonix.com
Joined 2003
Paid Member
Wow my phone butchered that... it's suppose to read "is just how bad the" not "is she told"....

The latest Samsung updates for my phone are ****. It doesn't just autospell, it auto changes words based on predictions.

Turn your predictive spell checking off - I've just had to do that for a relative who is visiting from overseas who has the new Samsung.

They seem much happier - no longer sending inappropriate texts . . .
 
Member
Joined 2016
Paid Member
I wouldn't trust fMRI here as far as I could throw a 7T machine. The field is rife with issues in terms of publishing utter garbage (and I really mean that as critically as I can) in studies on psychological or preference problems. Yes, certain sections of the brain light up with greater blood flow during these tests.

A neuroscientist I know once described using fMRI to map brain function as a bit like trying to work out how Windows works by analysing a PC with a thermal camera...
 
<snip>

Double blind and/or ABX ought not matter, or so my intuition tells me. At least if everything else is right. If there is some problem with any reasonable test methodology, somebody should figure out what it is, and then fix it. Or figure out an alternate methodology offering equally good experimental reliability.

The "everything else is right" part is the main difficulty ;)
We can be sure that usually something will be wrong, as humans are doing the experiments and some experimental decisions are just subjective.

But the main point should be the inclination of the experimenter to do a good test and to get correct results. Quite often imo the opposite seems to be true, people are using "blind" tests mainly as knockout argument, trying to "proof" audiophiles wrong are more than happy if their predefined believes were confirmed.

In addition, we probably can agree that some people who think they can hear small, subtle details, actually may not be very good at it.

Which of course is true......

Personally, I always get suspicious when people talk about things like "depth of sound field," etc., because I don't know what that means, or how to translate it to more well-defined technical terminology.

The tech doc EBU - 3286 or EBU 3286s1 are related to the subjective assessment of the quality of progamme material but it is nevertheless useful to use at a guide for the description of sound reproduction.

Or as Spikovski wrote in his technical review from 2000 (Assessment of Sound Field parameter differences in studio listening conditions):

"4.3. Imaging quality

The attributes relevant to interpretation of this factor relate to the room and the (phantom) sound sources which are simulated with the aid of stereophony in the loudspeaker base area. The attributes “stage width” and “stage depth” describe the imaging of the simulated sound source, e.g. an orchestra, relative to the dimensions “spatial breadth” and “depth”. Good spatial depth of the sound source is described with attributes such as “exactly localizable”, “transparent”, “detailed” and “spacious”, which are contrary to the attribute “blurred”. "

But, as stated earlier, the people doing the first stereophonic reproduction experiments in the 1930s already knew, that a listeners impression of depth of image will depend on experience in listening to real acoustic sources.

Of course to find the cause and effect relation in technical parameters of for example amplifiers will be much more complicated.....
 
While it's easy to agree that the last 40 years or so failed to significantly influence the overall quality of analog audio amplifiers, despite notable exceptions, it makes one wonder why is this so. Accumulated knowledge should have, by now, made at least many of even midrange priced amps sound better to much better - but it doesn't.

In my view, the primary reason for this is bean counting in manufacturers' practices, where shiny mostly useless LEDs are still preferred to function, like inceasing power supplies, output stage capacities, etc.

The secondary reason is that audio has altogether lost its once great importance to video and later on PC environments, these days phones and tablets. It simply is not as important today as it once was. We here are some of the last Mohicans.

The third reason is that the buying public has been devalued by DBTs, which bring us new "discoveries", "breakthroughs" and even "revolutions" month in, month out. New kids on the block have new toys to sell, so they need superlatives to sell, in name if not in fact.

Lastly, I think most users are actually too intimidated to trust their own hearing and say "No, I think product "X" is not worth the money" or some such. All the "revelations" pver the deacdes make them expect no less than sensations, each and every time. They are too afraid to admit even to themselves that they have been taken to the cleaners too many times, and have spent money on ads and gadgets far too often.

WE need to go back in time to the mid 70-ies and the times of the then eople like NAD who offered the 3020 integrated amp made of junk parts and hence cheap, but offered better value than just about everything else in plain sight. In that respect, John's idea of value better electronics is jsut what is needed, but it will be a VERY hard road to walk, there will be too many vested interest toes to walk over.
 
Member
Joined 2014
Paid Member
The problem with that is that you are missing the fact that considerable leaps have been made in cinema sound and headphone immersive soundfield generation. Sadly hamstrung by dolby licensing and vested interests, but progress nonetheless. It's just not been accepted by music lovers as worth the effort. So we argue minutiae over a flawed starting point (2 channel sound).
 
AX tech editor
Joined 2002
Paid Member
The problem with that is that you are missing the fact that considerable leaps have been made in cinema sound and headphone immersive soundfield generation. Sadly hamstrung by dolby licensing and vested interests, but progress nonetheless. It's just not been accepted by music lovers as worth the effort. So we argue minutiae over a flawed starting point (2 channel sound).

Maybe it is thought not worth the effort because stereo is so effective in duping us into thinking we hear 3D sound. It is a very credible illusion so you need to come up with something that is a LOT better in that respect to get people to pay the extra cost.

I think the reason why Dolby Atmos and Aura 3D and whatever they are called these days make headway into theathers and cinemas is because there it is not about the music per se but primarily about the effects. But for a pure audiophile it doesn't add much that is worthwhile.

Jan
 
Member
Joined 2014
Paid Member
I don't want 3D sound I want natural acoustic. Whenever I listen to live music I am aware of the space, which stereo rarely gets close to replicating. I have heard it once, but only once. I suspect the limited LF extension of my system has a part to play in this and I have a plan to fix that. But would still like to try out my (few) SACDs with full surround one day.
 
DBT are basically impossible by default. There's a null effect because as you increasingly add data points they're offset by the beers required to be provided to get someone to sit into the next step.

Reminds me of listening to a JC-80 with John Dennison, coupled with the fact that no matter how much you like Rickie Lee Jones how many times in a row can you listen to the first minute of an album.
 
Status
Not open for further replies.